Nonlinear Dynamical Friction from the Doppler-Shifted Equilibrium Memory Kernel

We present a statistical mechanics framework for modeling non-equilibrium transport coefficients using the Generalized Langevin Equation (GLE). We show that the kernel, obtained via the Fluctuation-Dissipation Theorem (FDT) from the stochastic force autocorrelation measured in a thermal equilibrium state, is sufficient to model the dynamics of the system in a Non-Equilibrium Steady State (NESS). This approach provides a computationally efficient path to modeling complex transport problems. We apply this framework to the canonical problem of test particle drag in a uniform plasma. The GLE formalism is shown to naturally capture non-Markovian phenomena through the moments of the kernel, including an effective mass renormalization and oscillatory relaxation. We demonstrate that the standard Chandrasekhar stopping power formula arises naturally as the Markovian limit of this equilibrium memory kernel. These theoretical predictions are quantitatively validated by direct Particle-in-Cell simulations, which confirm the predicted oscillatory structure of the memory kernel. This work thus establishes a practical method for predicting non-equilibrium transport properties from first-principles equilibrium simulations.

💡 Research Summary

The paper introduces a unified statistical‑mechanics framework for predicting nonlinear friction (stopping power) in non‑equilibrium steady states (NESS) using only equilibrium information. The authors start from the Mori‑Zwanzig projection formalism to derive a Generalized Langevin Equation (GLE) for a macroscopic variable A(t) coupled to a high‑dimensional thermal bath. In the GLE the memory kernel γ(t) encodes the retarded friction and, by the second Fluctuation‑Dissipation Theorem (FDT), is directly proportional to the force autocorrelation function of a particle held fixed in equilibrium.

A central theoretical advance is the “Doppler‑shifted FDT”. By considering a uniformly moving source ρ_ext(r,t)=Qδ(r−vt) in a linear medium (e.g., Vlasov‑Poisson plasma), the authors show that the Green’s function depends only on the lag τ through the combination x=vτ. In Fourier space this corresponds to a frequency shift ω→ω−k·v, i.e., a Doppler shift. Consequently, the non‑equilibrium friction coefficient ν(J) for a subsystem moving with constant flux J can be expressed as an integral over the equilibrium structure factor S_eq(k,ω) evaluated on the resonant manifold ω=k·J. Equation (11) encapsulates this result: the entire nonlinear transport curve is a kinematic sampling of the equilibrium spectrum.

To validate the theory, the authors apply it to the classic problem of a heavy ion traversing a collisionless electron‑proton plasma. They write the ion’s dynamics as a GLE (12) with memory kernel γ(ω) linked to the longitudinal dielectric function ε(k,ω) via Eq. (14). Using the Vlasov dielectric response, they obtain an explicit Gaussian‑shaped kernel (15). Substituting this kernel back into the GLE yields a non‑Markovian integro‑differential equation (16) for the mean velocity, which naturally includes transient buildup of the plasma wake and memory effects.

In the Markovian limit—when the ion velocity varies slowly compared to the plasma relaxation time—the convolution simplifies, and the drag force reduces to the real part of γ evaluated at the Doppler‑shifted frequency ω=k·v₀. This reproduces the Chandrasekhar stopping‑power formula (18‑19). The authors further expand the result in the low‑velocity (Stokes‑like linear drag, Eq. 20) and high‑velocity (1/v² decay, Eq. 21) limits, showing that a single equilibrium kernel captures both regimes.

Numerical validation is performed with high‑resolution 2D Particle‑in‑Cell (PIC) simulations using the EPOCH code. Two sets of simulations are executed: (A) equilibrium runs where “ghost” particles sample the electric‑field fluctuations without perturbing the plasma, allowing extraction of γ(t); (B) non‑equilibrium runs where a real test ion is launched at various speeds and its deceleration is measured. The drag curves obtained from set B match the predictions generated from the equilibrium kernel of set A with remarkable accuracy, including the oscillatory features of γ(t) that govern non‑Markovian behavior at high speeds.

The study demonstrates that nonlinear transport coefficients in NESS can be obtained from a single equilibrium measurement, reducing computational cost from O(N) separate non‑equilibrium simulations to O(1). This “Doppler‑shifted FDT” provides a practical tool for efficiently modeling transport in complex media such as dense plasmas, soft matter, and strongly correlated solids, where direct non‑equilibrium simulations are often prohibitive. The work also highlights the physical significance of the memory kernel’s structure—its oscillations, decay rates, and frequency dependence directly reflect underlying plasma processes like Landau damping and wake formation—offering new avenues for theoretical and experimental exploration.